【摘要】 神经网络在混沌系统控制中主要用于建模辨识或用作控制器,这2种作用都是以神经网络的任意非线性连续函数逼近能力为基础的.用神经网络进行建模辨识有2种途径:使用神经网络直接构造混沌系统的输入输出模型或通过神经网络构造混沌系统中的特征参量.神经网络应用于混沌系统建模辨识的研究主要集中在如何改进神经网络的学习算法和结构方面;神经网络用作控制器时主要用来解决混沌系统的轨迹跟踪或同步控制问题,其研究主要集中在改进控制结构与方法上.将神经网络应用于混沌系统控制已经取得了一些进展:动态神经网络开始应用于混沌系统辨识和控制;基于神经网络工作原理的函数网络能更容易地实现混沌系统的逼近.神经网络应用于混沌控制也存在一些需要解决的问题,如神经网络选择问题、神经网络结构问题、计算方法问题、硬件实现问题等.更多还原

【Abstract】 <Abstrcat>The uses of neural networks in chaos control can be concluded as follows:modeling or operating as controller.These uses are based on the capacity of neural networks that can approach any nonlinear continuous function.Modeling the chaos system by using neural networks can be realized in two ways.One is establishing the inputoutput model directly;the other is constructing the character parameters of chaos system.The research in this aspect is focused on improving the structure and arithmetic of neural networks,which have been applied in practice.The controller based on neural networks is usually used to solve the problems of following tracks or synchronization.The study in this aspect is focused on improving the structure and method of control.Some headway has been made;for example,dynamical neural networks have been applied and function networks have appeared which can approach chaos function easily.Some questions also exist.For example:how to choose the structure and arithmetic of neural networks and how to realize the function of neural networks by hardware.更多还原